© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it
    Neuronpedia
    Jacobian LensNEW
    Natural Language
    Autoencoders
    MAY 26    Assistant AxisJAN 26Circuit TracerAPR 26SteerSAE EvalsExportsinterp-engineNEW Community BlogPrivacy & TermsContact
    1. Home
    3. Gemma-2-27B
    5. 10-GEMMASCOPE-RES-131K
    7. 45377
    Prev
    Next
    INDEX
    Explanations

    constants ending in TYPE

    np_acts-logits-general · gemini-2.5-flash-lite

    The neuron is tuned to pick out the uppercase identifier “TYPE” (as seen in code macros and definitions).

    oai_token-act-pair · o4-miniTriggered by @jyhe0408
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    google/gemma-scope-27b-pt-res/layer_10/width_131k
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    No Configuration Found
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
    ↵↵
    -2.98
    ↵
    -2.89
    I
    -2.88
     letzte
    -2.73
     offizielle
    -2.61
     klassische
    -2.59
     vollständige
    -2.58
     We
    -2.58
     Preferably
    -2.58
     größte
    -2.55
    POSITIVE LOGITS
    黉
    3.42
    蹰
    3.22
     faſt
    3.20
    鋱
    2.95
    圌
    2.84
    嫪
    2.81
    硨
    2.81
    y
    2.78
    Ꮓ
    2.75
     juga
    2.70
    Activations Density 0.008%

    No Known Activations